Given the panel dataset of per capita gross regional domestic product (GRDP), I conduct HP filter technique to decompose it into trend and cyclical components. In this decomposition, do I need to confirm the data is stationary by unit-root test before filtering?
I suggest you to follow a thread of similar discussion here : https://www.researchgate.net/post/A_Question_about_VAR_estimation
For unit test, you may follow : http://campus.usal.es/~ehe/Papers/arranz.pdf
Thank you very much for your kind suggestions, Dr. Ali.
I have read your recommended reference. That is related with my question, but not addresses to it precisely. I should have described my question more precisely in line with my intention.
If possible, I would like to see any previous work that uses the unit-root test before the HP filtering or discussions, closely associated with my questions.
- Badges
- Science topic